It's an opportunity they grasped with both hands, says Alan Nahum, professor of physics at the Clatterbridge Centre for Oncology, UK, and lead organizer of the La Barcarolle workshop for the European Society for Therapeutic Radiology and Oncology (ESTRO). Here he gives his personal take on five of the big talking points to emerge during two days of group discussions.
The quest for success
In general, radiation treatment of cancer is highly successful, with some 80% of tumours "controlled" by the high radiation doses administered. However, given the high-tech equipment available and the precision of modern delivery methods, it is legitimate to look closer at that remaining 20% of cases.
With this in mind, there followed a wide-ranging discussion about how to learn from our radiotherapy treatment failures - e.g. to document exactly the location of the recurrence, be it in the gross tumour volume or outside it.
This firmed up into a collective desire to see detailed archiving systems set up to capture all of the essential information from a radiotherapy treatment: dose distributions overlaid on patient anatomy; fractionation schedule; dose-volume histograms; co-morbidity data; treatment follow-up data (local control or not?); early and late complications; information on the individual patient "biology" via blood samples, etc.
There was much discussion on how such an archive could be organized on a pan-European basis, including due recognition of the barriers currently in place (particularly tough in the UK) concerning the dissemination of any kind of patient data, whether anonymous or not.
It was noted by some of us that comprehensive patient databases based on treatment outcomes at our own clinics, together with powerful analysis software, would constitute an invaluable resource for improving radiotherapy (i.e. rather than some generalized outcome data culled from the literature).
From there, it would be straightforward enough to derive best-fit values of parameters in biological models (TCP, NTCP) that are specifically applicable to the local patient population. Software packages such as VODCA (which is purpose-built for the administration, organization and analysis of multicentre clinical trials in radiotherapy) are already available for this purpose.
To avoid duplication of effort, the consensus is that moves in this direction need to be coordinated with existing projects of a similar nature - e.g. GENEPI (Genetic pathways for the prediction of the effects of irradiation) and ESTRO-EQUAL (a quality-assurance network for photon and electron radiotherapy).
The merits of clinical trials
There was a robust debate about randomized clinical trials - more specifically, whether such trials are the only way to evaluate new technologies. The consensus seems to be that this bureaucratically cumbersome, resource-hungry mechanism isn't always essential (though to be fair, the radiation oncologists in the group defended the clinical-trial paradigm, emphasizing that uncontrolled confounding factors can bias results).
It was felt that physicists can model radiotherapy situations reasonably faithfully – this approach could be termed "the engineering paradigm". The key is to use the results we have to derive dose-effect relationships and then to apply these relationships to new situations. In my opinion, this is equivalent to using TCP and NTCP models to assess new treatment approaches.
External versus internal delivery
Another big talking point was external versus internal delivery of radiation therapy. The latter (also known as targeted therapy) is done via radioactive isotopes or unsealed sources which can, in principle, eliminate isolated metastases if the "targeting" is sufficiently specific. Currently, however, there are no software systems to enable the rational planning of combined external-beam and unsealed-source therapies.
This widened into a discussion of another type of combined therapy, one that is very much part of the clinical mainstream: chemotherapeutic agents administered together with radiotherapy. The consensus is that there's a pressing need to put such combined therapies on a rational, radiobiological basis. In order to do this, some fundamental experimental radiobiology is necessary - e.g. in-vitro quantitative studies of the cell-killing effect of combining these agents with ionizing radiation.
The imaging gap
The so-called "imaging gap" provided plenty of food for thought. Put simply, the lower detection limit for cutting-edge modalities such as PET and CT is of the order of 106 cells. And yet far fewer surviving tumour cells than this spells total treatment failure. At the other extreme, it is possible to use totally different histological techniques to image a single cell under the microscope.
How can this huge gap be bridged? It's felt that we need more detailed studies of the clonogen density (a clonogen is a cell that can proliferate and give rise to a colony of cells) in excised tissue samples - e.g. during surgery. Some estimate of clonogen number is essential if one is to make informed decisions about the desired dose level at the edge of, or just outside, the tumour. Currently we are groping in the dark.
Training harmonization
Two leading questions framed the discussion on professional development and training. Should there be a Europe-wide board-certification process, akin to the one in North America? And would such a process facilitate the movement of trained medical physicists between European countries?
There was much talk about harmonizing qualifications for medical physicists in the European Union and the near abroad. Currently, things are a mess - fragmentation reigns. It is virtually impossible to move from one country to another owing to a multiplicity of mutually incompatible "registration processes".
Organizations like the European Federation of Organizations for Medical Physics are working on this, but progress to date has been limited. The feeling among delegates is that this is not an insurmountable problem, but that no-one thus far has pushed hard enough for anything to change.